Focus max variability

[libraryman]

Hi, i guess that there are a lot of people out there using Focusmax...i have been using it for several months and iv'e run my V curves that appear o.k.

however the other night i ran into some odd behaviour that meant focus max failed for me.

i was using a 5nm narrowband filter.

Due to the fast temp drop; i started to record my focus position for the temp!  focusmax converged and gave a reading of position 22295....o.k.

 after some time i refocused on a different star..focus converged but it was clearly not best focus so several times i reverted to 22295 + a few.

i have struggled with the ideal brightness  for focusing through the 5NM filter.

is there a range of brightness that can be used for Narrowband?

does the choice of star make a big difference? i wasn't imaging at the zenith but not far off.

Is it wise to focus on the same star all night if possible?

Ray 

[perfrej]

I suggest you use the Acquirestar feature. It finds a suitable star that has no neighbors that can disturb the process. For NB I use mag 3-5, for LRGB mag 6-8. Make sure you use three exposures per measurement so that the seeing is cancelled out.

(The numbers from your focuser say very little when we do not know how many steps you have per FWHM value...)

/per

[libraryman]

Hi Per, i don't have the full version of Pin point, which i believe is required for Acquirestar,  though i accept it would be better since i could probably automate the slew control etc. 

The focuser moves 6.7 microns per/step ..since the Tak106 has a 55 micron critical zone that's 8 steps..however i have just measured the Tak focuser backlash and that's 22 steps or 147 microns.

it might be that's the issue...would you consider that possible?

Ray

[perfrej]

Yes it is possible, but it should be properly handled by focusmax. What kind of motor do you have on there?

/per

[libraryman]

Seletek, with a Seletek armadillo controller.

I did not have the backlash configured in focusmax either for the v curves i ran or for subsequent imaging runs.

However, i have now entered the backlash parameters I.e 22 steps.

Do you feel this is unnecessary?

Ray

[Martin-Devon]

Ray, for narrowband 5nM I usually focus in on a magnitude 4-5 star, and for broadband a 6-8 magnitude star. I prefer to chose the focus star either directly in (easy with broadband) my imaging field, or as close as possible (usually for narrowband) to my actual imaging field. It's simple to pick one of these stars on Stellarium or CdC, just scan for the closest star of the required magnitude, note it's RA/DEC, then do a slew/plate solve across to whenver you need to refocus, do the Focusmax routine, then slew back again - all very easy especially with automation such as SGP or CCD Commander, but easy also manually. The only times I've had issues with Focusmax giving misleading/wrong focus values is when there's been intermittent clouds hopping across or when I've mistakenly picked a double star for focusing. I've not used Acquirestar yet. 

For backlash, I have measured this for my Tak focuser with the Lakeside unit attached, and entered this figure in the Focusmax software. I have 55 clicks of backlash in my set-up, so my focuser has to move these 55 clicks before the focuser engages in terms of actual movement when it switches from going backwards to going forwards. Easy to measure,just put your finger on the other focuser knob on the other side to the Lakeside, and you'll easily feel when that knob is moving a click or not. 

If you've not tried it, the latest Focusmax 4 version seems to be better than the previous - although you now have to pay for this version 4 since it's marketed now by CCD Ware. 

PM me if you have any queries.

Martin 

[libraryman]

H Martin, V4 does look good but the price is high.

i suppose that 55 clicks sounds a larger backlash than mine but that depends on how far a step moves the focuser.

I have put the backlash setting in now so i'll check it next opportunity.

Ray